I want to encrypt large files or data streams

We recommend the Streaming AEAD primitive with AES128_GCM_HKDF_1MB key type for most file encryption use cases.

The Streaming Authenticated Encryption with Associated Data (Streaming AEAD) primitive is useful for encrypting live data streams or large files that don't fit in memory. Similar to AEAD, it is symmetric, using a single key for both encryption and decryption.

The following examples get you started using the Streaming AEAD primitive:

Go

import (
	"bytes"
	"fmt"
	"io"
	"log"
	"os"
	"path/filepath"

	"github.com/tink-crypto/tink-go/v2/insecurecleartextkeyset"
	"github.com/tink-crypto/tink-go/v2/keyset"
	"github.com/tink-crypto/tink-go/v2/streamingaead"
)

func Example() {
	// A keyset created with "tinkey create-keyset --key-template=AES256_CTR_HMAC_SHA256_1MB". Note
	// that this keyset has the secret key information in cleartext.
	jsonKeyset := `{
    "primaryKeyId": 1720777699,
    "key": [{
        "keyData": {
            "typeUrl": "type.googleapis.com/google.crypto.tink.AesCtrHmacStreamingKey",
            "keyMaterialType": "SYMMETRIC",
            "value": "Eg0IgCAQIBgDIgQIAxAgGiDtesd/4gCnQdTrh+AXodwpm2b6BFJkp043n+8mqx0YGw=="
        },
        "outputPrefixType": "RAW",
        "keyId": 1720777699,
        "status": "ENABLED"
    }]
	}`

	// Create a keyset handle from the cleartext keyset in the previous
	// step. The keyset handle provides abstract access to the underlying keyset to
	// limit the exposure of accessing the raw key material. WARNING: In practice,
	// it is unlikely you will want to use an insecurecleartextkeyset, as it implies
	// that your key material is passed in cleartext, which is a security risk.
	// Consider encrypting it with a remote key in Cloud KMS, AWS KMS or HashiCorp Vault.
	// See https://github.com/google/tink/blob/master/docs/GOLANG-HOWTO.md#storing-and-loading-existing-keysets.
	keysetHandle, err := insecurecleartextkeyset.Read(
		keyset.NewJSONReader(bytes.NewBufferString(jsonKeyset)))
	if err != nil {
		log.Fatal(err)
	}

	// Retrieve the StreamingAEAD primitive we want to use from the keyset handle.
	primitive, err := streamingaead.New(keysetHandle)
	if err != nil {
		log.Fatal(err)
	}

	// Create a file with the plaintext.
	dir, err := os.MkdirTemp("", "streamingaead")
	if err != nil {
		log.Fatal(err)
	}
	defer os.RemoveAll(dir)
	plaintextPath := filepath.Join(dir, "plaintext")
	if err := os.WriteFile(plaintextPath, []byte("this data needs to be encrypted"), 0666); err != nil {
		log.Fatal(err)
	}
	plaintextFile, err := os.Open(plaintextPath)
	if err != nil {
		log.Fatal(err)
	}

	// associatedData defines the context of the encryption. Here, we include the path of the
	// plaintext file.
	associatedData := []byte("associatedData for " + plaintextPath)

	// Encrypt the plaintext file and write the output to the ciphertext file. In this case the
	// primary key of the keyset will be used (which is also the only key in this example).
	ciphertextPath := filepath.Join(dir, "ciphertext")
	ciphertextFile, err := os.Create(ciphertextPath)
	if err != nil {
		log.Fatal(err)
	}
	w, err := primitive.NewEncryptingWriter(ciphertextFile, associatedData)
	if err != nil {
		log.Fatal(err)
	}
	if _, err := io.Copy(w, plaintextFile); err != nil {
		log.Fatal(err)
	}
	if err := w.Close(); err != nil {
		log.Fatal(err)
	}
	if err := ciphertextFile.Close(); err != nil {
		log.Fatal(err)
	}
	if err := plaintextFile.Close(); err != nil {
		log.Fatal(err)
	}

	// Decrypt the ciphertext file and write the output to the decrypted file. The
	// decryption finds the correct key in the keyset and decrypts the ciphertext.
	// If no key is found or decryption fails, it returns an error.
	ciphertextFile, err = os.Open(ciphertextPath)
	if err != nil {
		log.Fatal(err)
	}
	decryptedPath := filepath.Join(dir, "decrypted")
	decryptedFile, err := os.Create(decryptedPath)
	if err != nil {
		log.Fatal(err)
	}
	r, err := primitive.NewDecryptingReader(ciphertextFile, associatedData)
	if err != nil {
		log.Fatal(err)
	}
	if _, err := io.Copy(decryptedFile, r); err != nil {
		log.Fatal(err)
	}
	if err := decryptedFile.Close(); err != nil {
		log.Fatal(err)
	}
	if err := ciphertextFile.Close(); err != nil {
		log.Fatal(err)
	}

	// Print the content of the decrypted file.
	b, err := os.ReadFile(decryptedPath)
	if err != nil {
		log.Fatal(err)
	}
	fmt.Println(string(b))
	// Output: this data needs to be encrypted
}
streamingaead_test.go

Java

package streamingaead;

import static java.nio.charset.StandardCharsets.UTF_8;

import com.google.crypto.tink.InsecureSecretKeyAccess;
import com.google.crypto.tink.KeysetHandle;
import com.google.crypto.tink.RegistryConfiguration;
import com.google.crypto.tink.StreamingAead;
import com.google.crypto.tink.TinkJsonProtoKeysetFormat;
import com.google.crypto.tink.streamingaead.StreamingAeadConfig;
import java.io.IOException;
import java.nio.ByteBuffer;
import java.nio.channels.FileChannel;
import java.nio.channels.ReadableByteChannel;
import java.nio.channels.WritableByteChannel;
import java.nio.file.Files;
import java.nio.file.Path;
import java.nio.file.Paths;
import java.nio.file.StandardOpenOption;
import java.security.GeneralSecurityException;

/**
 * A command-line utility for encrypting files with Streaming AEAD.
 *
 * <p>It loads cleartext keys from disk - this is not recommended!
 *
 * <p>It requires the following arguments:
 *
 * <ul>
 *   <li>mode: Can be "encrypt" or "decrypt" to encrypt/decrypt the input to the output.
 *   <li>key-file: Read the key material from this file.
 *   <li>input-file: Read the input from this file.
 *   <li>output-file: Write the result to this file.
 *   <li>[optional] associated-data: Associated data used for the encryption or decryption.
 */
public final class StreamingAeadExample {
  private static final String MODE_ENCRYPT = "encrypt";
  private static final String MODE_DECRYPT = "decrypt";
  private static final int BLOCK_SIZE_IN_BYTES = 8 * 1024;

  public static void main(String[] args) throws Exception {
    if (args.length != 4 && args.length != 5) {
      System.err.printf("Expected 4 or 5 parameters, got %d\n", args.length);
      System.err.println(
          "Usage: java StreamingAeadExample encrypt/decrypt key-file input-file output-file"
              + " [associated-data]");
      System.exit(1);
    }
    String mode = args[0];
    Path keyFile = Paths.get(args[1]);
    Path inputFile = Paths.get(args[2]);
    Path outputFile = Paths.get(args[3]);
    byte[] associatedData = new byte[0];
    if (args.length == 5) {
      associatedData = args[4].getBytes(UTF_8);
    }

    // Initialize Tink: register all Streaming AEAD key types with the Tink runtime
    StreamingAeadConfig.register();

    // Read the keyset into a KeysetHandle
    KeysetHandle handle =
        TinkJsonProtoKeysetFormat.parseKeyset(
            new String(Files.readAllBytes(keyFile), UTF_8), InsecureSecretKeyAccess.get());

    // Get the primitive
    StreamingAead streamingAead =
        handle.getPrimitive(RegistryConfiguration.get(), StreamingAead.class);

    // Use the primitive to encrypt/decrypt files
    if (mode.equals(MODE_ENCRYPT)) {
      encryptFile(streamingAead, inputFile, outputFile, associatedData);
    } else if (mode.equals(MODE_DECRYPT)) {
      decryptFile(streamingAead, inputFile, outputFile, associatedData);
    } else {
      System.err.println(
          "The first argument must be either "
              + MODE_ENCRYPT
              + " or "
              + MODE_DECRYPT
              + ", got: "
              + mode);
      System.exit(1);
    }
  }

  private static void encryptFile(
      StreamingAead streamingAead, Path inputFile, Path outputFile, byte[] associatedData)
      throws GeneralSecurityException, IOException {
    try (WritableByteChannel encryptingChannel =
            streamingAead.newEncryptingChannel(
                FileChannel.open(outputFile, StandardOpenOption.WRITE, StandardOpenOption.CREATE),
                associatedData);
        FileChannel inputChannel = FileChannel.open(inputFile, StandardOpenOption.READ)) {
      ByteBuffer byteBuffer = ByteBuffer.allocate(BLOCK_SIZE_IN_BYTES);
      while (true) {
        int read = inputChannel.read(byteBuffer);
        if (read <= 0) {
          return;
        }
        byteBuffer.flip();
        while (byteBuffer.hasRemaining()) {
          encryptingChannel.write(byteBuffer);
        }
        byteBuffer.clear();
      }
    }
  }

  private static void decryptFile(
      StreamingAead streamingAead, Path inputFile, Path outputFile, byte[] associatedData)
      throws GeneralSecurityException, IOException {
    try (ReadableByteChannel decryptingChannel =
            streamingAead.newDecryptingChannel(
                FileChannel.open(inputFile, StandardOpenOption.READ), associatedData);
        FileChannel outputChannel =
            FileChannel.open(outputFile, StandardOpenOption.WRITE, StandardOpenOption.CREATE)) {
      ByteBuffer byteBuffer = ByteBuffer.allocate(BLOCK_SIZE_IN_BYTES);
      while (true) {
        int read = decryptingChannel.read(byteBuffer);
        if (read <= 0) {
          return;
        }
        byteBuffer.flip();
        while (byteBuffer.hasRemaining()) {
          outputChannel.write(byteBuffer);
        }
        byteBuffer.clear();
      }
    }
  }

  private StreamingAeadExample() {}
}
StreamingAeadExample.java

Python

"""A command-line utility for using streaming AEAD for a file.

It loads cleartext keys from disk - this is not recommended!

It requires 4 arguments (and one optional one):
  mode: either 'encrypt' or 'decrypt'
  keyset_path: name of the file with the keyset to be used for encryption or
    decryption
  input_path: name of the file with the input data to be encrypted or decrypted
  output_path: name of the file to write the ciphertext respectively plaintext
    to
  [optional] associated_data: the associated data used for encryption/decryption
    provided as a string.
"""

from typing import BinaryIO

from absl import app
from absl import flags
from absl import logging
import tink
from tink import secret_key_access
from tink import streaming_aead

FLAGS = flags.FLAGS
BLOCK_SIZE = 1024 * 1024  # The CLI tool will read/write at most 1 MB at once.

flags.DEFINE_enum('mode', None, ['encrypt', 'decrypt'],
                  'Selects if the file should be encrypted or decrypted.')
flags.DEFINE_string('keyset_path', None,
                    'Path to the keyset used for encryption or decryption.')
flags.DEFINE_string('input_path', None, 'Path to the input file.')
flags.DEFINE_string('output_path', None, 'Path to the output file.')
flags.DEFINE_string('associated_data', None,
                    'Associated data used for the encryption or decryption.')


def read_as_blocks(file: BinaryIO):
  """Generator function to read from a file BLOCK_SIZE bytes.

  Args:
    file: The file object to read from.

  Yields:
    Returns up to BLOCK_SIZE bytes from the file.
  """
  while True:
    data = file.read(BLOCK_SIZE)
    # If file was opened in rawIO, EOF is only reached when b'' is returned.
    # pylint: disable=g-explicit-bool-comparison
    if data == b'':
      break
    # pylint: enable=g-explicit-bool-comparison
    yield data


def encrypt_file(input_file: BinaryIO, output_file: BinaryIO,
                 associated_data: bytes,
                 primitive: streaming_aead.StreamingAead):
  """Encrypts a file with the given streaming AEAD primitive.

  Args:
    input_file: File to read from.
    output_file: File to write to.
    associated_data: Associated data provided for the AEAD.
    primitive: The streaming AEAD primitive used for encryption.
  """
  with primitive.new_encrypting_stream(output_file,
                                       associated_data) as enc_stream:
    for data_block in read_as_blocks(input_file):
      enc_stream.write(data_block)


def decrypt_file(input_file: BinaryIO, output_file: BinaryIO,
                 associated_data: bytes,
                 primitive: streaming_aead.StreamingAead):
  """Decrypts a file with the given streaming AEAD primitive.

  This function will cause the program to exit with 1 if the decryption fails.

  Args:
    input_file: File to read from.
    output_file: File to write to.
    associated_data: Associated data provided for the AEAD.
    primitive: The streaming AEAD primitive used for decryption.
  """
  try:
    with primitive.new_decrypting_stream(input_file,
                                         associated_data) as dec_stream:
      for data_block in read_as_blocks(dec_stream):
        output_file.write(data_block)
  except tink.TinkError as e:
    logging.exception('Error decrypting ciphertext: %s', e)
    exit(1)


def main(argv):
  del argv

  associated_data = b'' if not FLAGS.associated_data else bytes(
      FLAGS.associated_data, 'utf-8')

  # Initialise Tink.
  try:
    streaming_aead.register()
  except tink.TinkError as e:
    logging.exception('Error initialising Tink: %s', e)
    return 1

  # Read the keyset into a keyset_handle.
  with open(FLAGS.keyset_path, 'rt') as keyset_file:
    try:
      text = keyset_file.read()
      keyset_handle = tink.json_proto_keyset_format.parse(
          text, secret_key_access.TOKEN
      )
    except tink.TinkError as e:
      logging.exception('Error reading key: %s', e)
      return 1

  # Get the primitive.
  try:
    streaming_aead_primitive = keyset_handle.primitive(
        streaming_aead.StreamingAead)
  except tink.TinkError as e:
    logging.exception('Error creating streaming AEAD primitive from keyset: %s',
                      e)
    return 1

  # Encrypt or decrypt the file.
  with open(FLAGS.input_path, 'rb') as input_file:
    with open(FLAGS.output_path, 'wb') as output_file:
      if FLAGS.mode == 'encrypt':
        encrypt_file(input_file, output_file, associated_data,
                     streaming_aead_primitive)
      elif FLAGS.mode == 'decrypt':
        decrypt_file(input_file, output_file, associated_data,
                     streaming_aead_primitive)


if __name__ == '__main__':
  flags.mark_flag_as_required('mode')
  flags.mark_flag_as_required('keyset_path')
  flags.mark_flag_as_required('input_path')
  flags.mark_flag_as_required('output_path')
  app.run(main)
streaming_aead_cli.py

Streaming AEAD

The Streaming AEAD primitive provides authenticated encryption for streaming data. It is useful when the data to be encrypted is too large to be processed in a single step. Typical use cases include encryption of large files or live data streams.

Encryption is done in segments, which are bound to their location within a ciphertext and cannot be removed or reordered. Segments from one ciphertext cannot be inserted into another ciphertext. To modify an existing ciphertext, the entire data stream must be re-encrypted.¹

Decryption is fast because only a portion of the ciphertext is decrypted and authenticated at a time. Partial plaintexts are obtainable without processing the entire ciphertext.

Streaming AEAD implementations fulfill the AEAD definition and are nOAE-secure. They have the following properties:

• Secrecy: Nothing about the plaintext is known, except its length.
• Authenticity: It is impossible to change the encrypted plaintext underlying the ciphertext without being detected.
• Symmetric: Encrypting the plaintext and decrypting the ciphertext is done with the same key.
• Randomization: Encryption is randomized. Two messages with the same plaintext yield different ciphertexts. Attackers cannot know which ciphertext corresponds to a given plaintext.

Associated data

Caution: Associated data is authenticated but NOT encrypted.

The Streaming AEAD primitive can be used to tie ciphertext to specific associated data. Suppose you have a database with the fields user-id and encrypted-medical-history: In this scenario, user-id can be used as associated data when encrypting encrypted-medical-history. This prevents an attacker from moving medical history from one user to another.

Choose a key type

We recommend AES128_GCM_HKDF_1MB for most uses. Generally:

• AES-GCM-HKDF
  • AES128_GCM_HKDF_1MB (or AES256_GCM_HKDF_1MB) is the faster option. It can encrypt 2⁶⁴ files with up to 2⁶⁴ bytes each. ~1 MB of memory is consumed during the encryption and decryption process.
  • AES128_GCM_HKDF_4KB consumes ~4 KB of memory and is a good choice if your system doesn't have a lot of memory.
• AES-CTR HMAC
  • AES128_CTR_HMAC_SHA256_1MB (or AES256_CTR_HMAC_SHA256_1MB) is a more conservative option.

Note: For 1 MB schemes, the plaintext may have any length within 0 to 2⁵¹ bytes.²

Security guarantees

Streaming AEAD implementations offer:

• CCA2 security.
• At least 80-bit authentication strength.
• The ability to encrypt at least 2⁶⁴ messages³ with a total of 2⁵¹ bytes² . No attack with up to 2³² chosen plaintexts or chosen ciphertexts has a probability of success larger than 2^-32.

Caution: Streaming AEAD offers no secrecy guarantees for associated data.

---

1. A reason for this restriction is the use of the AES-GCM cipher. Encrypting a different plaintext segment at the same location would be equivalent to reusing the IV, which violates the security guarantees of AES-GCM. Another reason is that this prevents roll-back attacks, where the attacker may try to restore a previous version of the file without detection. ↩

2. 2³² segments are supported, with each segment containing segment_size - tag_size bytes of plaintext. For 1 MB segments, the total plaintext size is 2³² * (2²⁰-16) ~= 2⁵¹ bytes. ↩

3. Streaming AEAD becomes insecure when a derived key (128-bit) and nonce prefix (independent random 7-byte value) combination is repeated. We have 184-bit collision resistance, which roughly translates to 2⁶⁴ messages if we want success probability to be less than 2^-32. ↩